[helene_t]

Fluffy, on Oct 30 2007, 01:59 PM, said:

I'd had some theological problem, how many boxes do I need to open to consider opening one more suicidal?

9999. What has this to do with theology?

[cherdano]

There is a room with 10,000 boxes. In one of them is a bomb. One box is picked at random. You have the choice of paying 1000$, or you will die if the box contains the bomb...

[jtfanclub]

Jlall, on Oct 29 2007, 09:39 PM, said:

if the boxes were only 1000 dollars I would not open any. To get anything significant I would need to put up far too much risk, and if my risk is very small then so is my gain so it's still not worth it.

1/10,000 chance of getting killed?


Let's see. You have about a 1.5 in 1 million chance of getting killed for every mile you drive. So, unless my math is wrong, you have at least a 1/10,000 chance of getting killed every time you take a car trip of 70 miles or more.

So let's put it a different way. How far would you be willing to drive to get a thousand bucks?

I'd probably open half the boxes if I could be sure that my next of kin would get the money. Five million smackers on the average, huh? I hope that they'd appreciate it.

[helene_t]

Right, it seems as if I wouldn't accept a $1000-offer for wearing a helmet during one of my daily 10-minute bicycle trips.

Difficult to be consistent when it comes to risk management. During the BSE crisis, there were smokers who didn't dare to eat beef. How much beef does it take to expose yourself to the health risk of a single cigarette? As I recall it was about 500 kg.

[cherdano]

jtfanclub, on Oct 30 2007, 09:36 AM, said:

Jlall, on Oct 29 2007, 09:39 PM, said:

if the boxes were only 1000 dollars I would not open any. To get anything significant I would need to put up far too much risk, and if my risk is very small then so is my gain so it's still not worth it.

1/10,000 chance of getting killed?


Let's see. You have about a 1.5 in 1 million chance of getting killed for every mile you drive. So, unless my math is wrong, you have at least a 1/10,000 chance of getting killed every time you take a car trip of 70 miles or more.

So let's put it a different way. How far would you be willing to drive to get a thousand bucks?

I'd probably open half the boxes if I could be sure that my next of kin would get the money. Five million smackers on the average, huh? I hope that they'd appreciate it.

These numbers can't be right. A 1/100 chance of getting killed in a year of driving 7,000 miles?

[jtfanclub]

cherdano, on Oct 30 2007, 10:46 AM, said:

These numbers can't be right. A 1/100 chance of getting killed in a year of driving 7,000 miles?

It's not quite additive: your odds are dying on mile 1 are 1.5/1 million, your odds of dying on mile 2 are slightly less (since you may have died on mile 1).

But I may certainly have screwed up the math. Here's the site with the 1.48 per million miles traveled:

http://www.usatoday.com/money/autos/2004-0...ic-deaths_x.htm

[bid_em_up]

"Tougher drunken-driving laws, stricter seat-belt enforcement and safer cars combined to reduce the number of people killed in car crashes to 1.48 for every 100 million miles traveled in 2003"

This is not the same as saying any one individuals odds of being killed in car crash are 1 in 1.48 million. (And that's not even considering that it says 100 million miles, not one million).

According to the National Safety Council the odds are 1 in 6,498 that you will die in a motor vehicle accident (in the course of one year) and 1 in 84 over the course of your lifetime.

But, "One year odds are approximated by dividing the 2003 population (290,850,005) by the number of deaths."

http://www.nsc.org/l...atinfo/odds.htm

Of course, this same info lists your odds of dying by legal execuation as:

1 in 4,847,500 in the course of the next year, and 1 in 62,468 over the course of a life time, while the real odds of it actually occuring are probably actually nil. Unless, of course, you are planning to commit some kind of capital crime in the near future (and then only if done in a state that has the death penalty).

Your individual odds would vary based on your driving habits, number miles travelled, etc.

[matmat]

bid_em_up, on Oct 30 2007, 12:33 PM, said:

According to the National Safety Council the odds are 1 in 6,498 that you will die in a motor vehicle accident (in the couse of one year) and 1 in 84 over the course of your lifetime.

does that mean that ~1.3% of deaths are vehicular?

[jtfanclub]

matmat, on Oct 30 2007, 12:43 PM, said:

bid_em_up, on Oct 30 2007, 12:33 PM, said:

According to the National Safety Council the odds are 1 in 6,498 that you will die in a motor vehicle accident (in the couse of one year)  and 1 in 84 over the course of your lifetime.

does that mean that ~1.3% of deaths are vehicular?

Actually, it's 1.7%. Dunno why it's higher- pedestrians, maybe?

Bid is right, it's per 100 million, not per 1 million. So it would take a 7,000 mile drive to be equal to 1/100 of 1%. With the usual caveats, not every driver is equally stupid, etc.

[bid_em_up]

matmat, on Oct 30 2007, 12:43 PM, said:

bid_em_up, on Oct 30 2007, 12:33 PM, said:

According to the National Safety Council the odds are 1 in 6,498 that you will die in a motor vehicle accident (in the couse of one year)  and 1 in 84 over the course of your lifetime.

does that mean that ~1.3% of deaths are vehicular?

No. Yes. I don't know.

The NSC's report considers a lot of things to be "motor vehicle accidents"

Animal rider or occupant of animal-drawn vehicle
Bus Occupant
Car Occupant
Air and Space Transport accidents
Construction Equipment
Pedestrians

etc.

There does not appear to be a category specifically for "drivers of automobiles".

[keylime]

My view is that, since I can't take it with me, I'm not going to be that interested in taking a risk that at age 31, my life ends because of heightened greed. Even if I sorely needed the money, life is just too precious and is meant to be lived. It is too vague and non-equivalent of an argument to assert that because I'm bypassing money, that I am not being philosophically adherent when I hop on the Metro to go into work or drive the 4 hours to Pittsburgh.

Greed is a recipe for bad things to happen.

[Echognome]

How about a couple of variants that are less severe than dying?

Game 1:
You have 10 boxes in a room, nine of them with $10,000 in them. However, one of them is the "Bankrupt" box and gives you nothing. How many boxes would you open?

Game 2:
Same set-up. Only 10 boxes, but nine of them are "prizes" and one of them is "Bankrupt". For opening up the first "prize" box, you get $1000. For each subsequent successful "prize" box you open you double your money. However, if you ever open the "Bankrupt" box, you get nothing. How many boxes would you open?

Which game would you rather play?

[Mbodell]

I'd much rather play game 2. I'd open 8 boxes and have an EV of $51,000. Much better than open 5 boxes with an EV of $25,000 from game 1. The down side here is much better than death and the stakes aren't so super great, especially when you phrase it in the win fall model.

[helene_t]

Game 1: When I have opened four boxes, the value of opening a fifth is 5/6 * 10000 - 1/6 * 40000, this is just positive. The sixth box I would not open. This gives me a 50% chance of EUR 50000, or 25000 on average.

Game 2: As long as the probability of doubling is higher than bankrupt, I would go on. Say 8 boxes. This gives me a 20% chance of winning EUR 128000, or 25600 on average.

[Echognome]

So would you rather have a coin flip for $50,000 or a 20% chance at $128,000? The EV for the latter is slightly greater.

I modelled the games adding some risk aversion and it doesn't take a whole lot to make the first game better, despite the second game having a higher EV.

[barmar]

zasanya, on Oct 30 2007, 04:51 AM, said:

I will open the 1st box the day I become convinced that there is no other way for me to earn money.I don't think that day would ever come,touch wood.

That seems to suppose that this game is played as an alternative to earning money some other way. But this game is presumably played during your leisure time, not in place of your job or other money-making activities. It's earned in addition to anything else you might earn, so I don't think it makes sense to compare it against other ventures.

Furthermore, the nature of thought experiments like this is that they're a "closed universe". Your only choices are to open boxes or not open boxes -- there's no third choice of "play rubber bridge for $100/point." That would be a different game.

[P_Marlowe]

Echognome, on Oct 30 2007, 07:03 PM, said:

So would you rather have a coin flip for $50,000 or a 20% chance at $128,000? The EV for the latter is slightly greater.

I modelled the games adding some risk aversion and it doesn't take a whole lot to make the first game better, despite the second game having a higher EV.

The coin flip, and it is not even close.

While the EV in the other game is slightly
higher, I will more often win money with
the coin flip.

With kind regards
Marlowe

[han]

You willl more often win money if you open only one box.

I would go for the coin flip as well btw.